The extraction of olfactory evoked potentials in human brain using independent component analysis - Based approach

Patients suffering from olfaction impairment are incapable of judging the smell of delicious food and fragrance. Their lack of smell vigilance might also make them unaware of danger environments, e.g., the leakage of carbon-oxide (CO). Traditional examination in clinics uses Sniffin sticks, which store odorants in sticks, and requests subjects to subjectively report what smell they have sensed. Nevertheless, subject's subjective feeling may inevitably bias the examination results. Accordingly, a objective method for olfactory examinations in otolaryngology clinics is needed. In this study, we tried to develop an objective method on magnetoencephalography (MEG) for olfaction examination by extraction olfactory evoked fields (OEF). Owing to the high noise level in MEG data, independent component analysis (ICA) was adopted to removal OEF-unrelated noises. ICA can decompose the measured EEG data into different independent components (IC). ICs were clustered into different groups based on their spatial weight over MEG channels, and only those ICs with spatial mapping closed to second somatosensory area (SII) were utilized for signal reconstruction. The source locations of extracted OEF signals were superimposed on individual anatomical MRI for display purpose. The proposed method may provide an efficacious way to enhance the SNR of OEFs.